(57 days)
The SIGNA PET/MR system combines magnetic resonance diagnostic devices (MRDD) and Positron Emission Tomography (PET) scanners that provide registration and fusion of high resolution physiologic and anatomic information, acquired simultaneously and isocentrically. The combined system maintains independent functionality of the MR and PET devices, allowing for single modality MR and / or PET imaging.
These systems are intended to be utilized by appropriately trained health care professionals to aid in the detection, localization, and diagnosis of diseases and disorders. MR is intended to produce transverse, sagittal, coronal and oblique cross-sectional MR images, spectroscopic images and/or spectra, and displays the internal structure and/or function of the human body. Other physical parameters derived from the images and/or spectra may also be produced. Depending on the region of interest, approved contrast agents may be used, as described in their labeling. This system may also be used for imaging during interventional procedures when performed with MR compatible devices, such as MR safe biopsy needles.
PET images and measures the distribution of PET radiopharmaceuticals in humans to aid the physician in determining various metabolic (molecular) and physiologic functions within the human body for evaluation of diseases and disorders such as, but not limited to, cardiovascular disease, neurological disorders and cancer.
The combined system utilizes the MR for radiation correction maps for PET studies. The system provides inherent anatomical reference for the fused PET and MR images due to precisely aligned MR and PET image coordinate systems.
The GE SIGNA PET/MR system is a combined Magnetic Resonance Diagnostic Device (MRDD) and Positron Emission Tomography (PET) scanner. The system is designed for whole body oncology, neurology and cardiology examinations. The SIGNA PET/MR system provides simultaneous acquisition of high resolution metabolic and anatomic information from the two major components of each system (MR and PET). Additional components of the system include: a detachable patient table and both the acquisition and processing workstations with associated software.
The SIGNA PET/MR includes a 3.0T superconducting magnet, gradient coil and a transmit/receive whole body radiofrequency coil. The system includes patient adaptable RF shimming capabilities. The SIGNA PET detectors are integrated into the MR bore. This allows for simultaneous, precisely aligned whole body MR and PET acquisitions. The PET subsystem supports Time of Flight (ToF) coincidence detection. The SIGNA PET/MR software is used for patient management, data management, scan control, image reconstruction and image archival and evaluation. All images conform to DICOM imaging format requirements.
The modifications to this system include the MotionFree Brain software feature, which allows users the flexibility to correct patient head motion using the acquired PET data from the exam, without the need of external tracking devices, additional MR data, or other motion tracking data schemes.
The provided text describes the 510(k) summary for the GE SIGNA PET/MR system with a specific modification: the "MotionFree Brain" software feature. This feature aims to correct patient head motion in PET data without external tracking devices or additional MR data.
Here's an analysis of the acceptance criteria and the study proving the device meets them, based on the provided text:
1. Table of Acceptance Criteria and Reported Device Performance
The document does not explicitly present a table of acceptance criteria with corresponding performance metrics. However, it states that "Performance testing on phantoms as part of non-clinical testing demonstrated MotionFree Brain achieved performance claims for Quantitation, Temporal Resolution, and Spatial Accuracy." For clinical testing, the acceptance was based on reader preference and confirmation that the feature could be "used safely and effectively in a clinical setting."
Implicit Acceptance Criteria and Reported Performance (derived from text):
| Performance Metric | Acceptance Criteria (Implicit) | Reported Device Performance |
|---|---|---|
| Quantitation (Phantom) | Performance claims achieved. | Achieved performance claims for Quantitation. |
| Temporal Resolution (Phantom) | Performance claims achieved. | Achieved performance claims for Temporal Resolution. |
| Spatial Accuracy (Phantom) | Performance claims achieved. | Achieved performance claims for Spatial Accuracy. |
| Clinical Efficacy/Safety | Device can be used safely and effectively in a clinical setting; no new questions of safety and effectiveness were raised; performs as well as or better than the predicate device. | "Clinical testing confirms that MotionFree Brain can be used safely and effectively in a clinical setting." "No new questions of safety and effectiveness were raised during nonclinical testing." "Testing demonstrated that the device is as safe, as effective, and performs as well as or better than the predicate device." Readers were asked to report their preference, implying that a favorable preference was the goal. |
| Substantial Equivalence | Software feature is as safe and effective as the predicate, and does not raise different questions of safety and effectiveness. | "The SIGNA PET/MR with the modified software feature has the same intended use as the predicate... performance data demonstrating that the feature is as safe and effective as the predicate, and does not raise different questions of safety and effectiveness." |
2. Sample Size Used for the Test Set and Data Provenance
- Test Set Sample Size: The document does not explicitly state the numerical sample size (number of cases or patients) used for the clinical test set. It mentions "randomly labeled cases."
- Data Provenance: The document does not specify the country of origin for the data. The study was described as a "retrospective blind study."
3. Number of Experts Used to Establish the Ground Truth for the Test Set and Qualifications
- Number of Experts: The document refers to "board-certified readers" (plural), indicating more than one expert. The exact number is not provided.
- Qualifications of Experts: The experts were "board-certified readers." No further details on years of experience or specific specializations (e.g., neuroradiologist) are given.
4. Adjudication Method for the Test Set
The document states: "The readers were blinded to feature use (e.g. whether feature was enabled or disabled), report case history, as well as to the assessments made by the other readers. The readers were asked to complete an assessment, including additional commentary, and report their preference on the pair of image series presented."
This describes a blinded read comparison where readers evaluated paired images (with and without the feature). The adjudication method primarily involved readers reporting their preference rather than a formal consensus or 2+1/3+1 type of adjudication. The text implies individual reader assessment and preference reporting, not a group adjudication to establish a "ground truth" through consensus for each case regarding motion correction per se.
5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study and Effect Size
- MRMC Study: An "external reader evaluation study was performed for MotionFree Brain. The retrospective blind study involved board-certified readers who were asked to evaluate randomly labeled cases that were reconstructed with and without MotionFree Brain." This structure implies an MRMC design, where multiple readers evaluate multiple cases under different conditions (with/without AI assistance).
- Effect Size of Human Readers Improvement: The document does not quantify the effect size (e.g., AUC, sensitivity, specificity improvement) of how much human readers improved with AI (MotionFree Brain) assistance vs. without. It focuses on reader preference and overall safety/effectiveness conclusion.
6. Standalone (Algorithm Only Without Human-in-the-Loop Performance) Study
The document describes "Performance testing on phantoms" for "Quantitation, Temporal Resolution, and Spatial Accuracy." This can be considered a form of standalone performance evaluation for the algorithm's core functionality, as it assesses the algorithm's output (reconstructed image quality metrics) on controlled phantom data, independent of human interpretation for diagnostic tasks.
7. Type of Ground Truth Used
- For Phantom Testing: The ground truth was based on the known, controlled characteristics of the phantoms used, against which the "Quantitation, Temporal Resolution, and Spatial Accuracy" were measured.
- For Clinical/Reader Study: The "ground truth" was indirectly established through the comparative assessment and preference of experienced, board-certified readers. It's not a definitive clinical outcome (e.g., pathology, long-term follow-up) but rather a relative assessment of image quality and diagnostic confidence as perceived by experts when comparing images with and without motion correction. The study design focused on assessing the impact of the feature on the interpretability by clinicians, rather than establishing a true disease status for each patient.
8. Sample Size for the Training Set
The document does not provide any information about the training set size for the MotionFree Brain software feature. This information is typically proprietary and not included in a 510(k) summary unless specifically requested or deemed critical for demonstrating substantial equivalence. Given that the feature "derives head motion information from the PET data" and "measures and incorporates the rigid-body motion information into the PET reconstruction," it likely involves algorithmic processing rather than a purely deep learning approach requiring a massive labeled training set in the conventional sense, though validation or tuning would still involve data.
9. How the Ground Truth for the Training Set Was Established
Since no information on the training set (or its existence as a distinct "training set" in a machine learning sense) is provided, there is no description of how its ground truth was established. If the algorithm is rule-based or model-based rather than solely data-driven (e.g., deep learning), a "training set" with ground truth in the supervised learning sense might not be applicable. The description ("MotionFree Brain derives head motion information from the PET data... measures and incorporates the rigid-body motion information into the PET reconstruction, correcting the position of each coincidence event") suggests a more deterministic or model-based approach to motion correction.
§ 892.1200 Emission computed tomography system.
(a)
Identification. An emission computed tomography system is a device intended to detect the location and distribution of gamma ray- and positron-emitting radionuclides in the body and produce cross-sectional images through computer reconstruction of the data. This generic type of device may include signal analysis and display equipment, patient and equipment supports, radionuclide anatomical markers, component parts, and accessories.(b)
Classification. Class II.